In cardiology, heart disease is the leading cause of death in humans; it encompasses conditions such as CAD, myocardial infection, and arrhythmia. Early detection of disease with its risk stratification, which are reducing rate of morbidity and mortality. In the conventional approach, the disease is diagnosed based on clinical assessment, ECG, EEG, CT and other laboratory investigations. Heart disease is a major concern in humans. It causes silent deaths in humans with a small pain in the chest area. In the early detection of disease, the clinical parameters of the patient. Before operating on the patient, it is necessary to stress test or an angiography on the patient according to the patient's history. In this study for heart disease detection with their perception in early stages, various deep learning models are used, such as LSTM and DNN, based on demographic and biochemical parameters in this proposed model. Various results improved the detection in heart disease by using the classification evaluation metrics. In this finding the deep learning models support the risk assessment in heart disease. LSTM obtained the better results in term of accuracy that achieved through this model.

TBA..